Hi BaiYang, Thanks for your reply!
For your question, my answer is below: I don’t need to care about Kafka brokers because I use AWS’s MSK product. For producers, I didn’t restart it, which is a cluster services and provided by other team.Consumers will be restarted when my application is redeployed. Thanks， Pin > On Apr 1, 2022, at 23:10, BaiYang <by_2...@163.com> wrote: > > how about stop the kafka brokers and restart them and restart producers and > consumers? > > > > > ---- 回复的原邮件 ---- > | 发件人 | Xiong, Pin<pinxion...@foxmail.com> | > | 日期 | 2022年03月31日 22:14 | > | 收件人 | users@kafka.apache.org<users@kafka.apache.org> | > | 抄送至 | | > | 主题 | [Question] What is the best practice in consuming multiple topics? | > To whom it may concern, > > I have a very different problem when using Kafka, and hope someone can help > me! > > Background > > I need to consume several different topics in my java project. For each > topic, they has different configurations, such as max.pull.records, but the > same consumer group. > > Also, there are different business logic for each topic. > > > How did I do > > I have created different KafkaConsumer instances with different topics. > Because I can setup the configurations only when creating the KafkaConsumer > instance. > > > Problem > > Some topic partitions cannot be consumed. This is really a mazing phenomenon! > I found that the client.id will be changed when I redeployed and restarted > the project. > > All of KafkaConsumer share with the global static variable > CONSUMER_CLIENT_ID_SEQUENCE in ConsumerConfig if I didn’t set client.id . > > Is it the best practice to configure the different client.id for each > KafkaConsumer instance? Or who can tell me how to solve my issue? > > > Looking forward to your reply! > > Thanks， > Pin >
